I have created a group of Linux SNMP monitors and assigned them to the LINUX SNMP hostgroup. Any server in that group will pick these monitors up. Now I want to setup a Service Escalation for a few of the monitors that are targeted to the LINUX SNMP hostgroup. When I go into the NagiosXI Core Config Manager to create a Service Escalation for the specific snmp monitors I believe are critical I can't get any of them to display. Am I doing something wrong? When I have a monitor targeted specifically to a host it will come up in the services box. If I remove all but 1 of the hosts from the hostgroup I can get a list of monitors also. I'm only having an issue with monitors that are targeted to a hostgroup that has more then 1 host targeted to it.
Please Help!
Thank you
Service Escalation lists no services
Re: Service Escalation lists no services
I believe this is a bug in the XI interface and not nagios core. I removed all Linux hosts but 1 in my linux snmp hostgroup. I then created a service escalation targeted to this hostgroup, saw all the services and selected *. After the escalation was created I went back to the linux snmp hostgroup and re-added all the hosts back. I verified the escalation was working. I then went back into the service escalation to make an edit and wasn't able to do so. The only way for me to edit the service escalation is to remove all but 1 host from the hostgroup again. This is a very troublesome bug if you build out your monitoring solution to have services automatically targeted to hosts in a hostgroup.
Can anyone think of an easy way to add a configuration when the GUI isn't working correctly?
Thank you
Can anyone think of an easy way to add a configuration when the GUI isn't working correctly?
Thank you
Re: Service Escalation lists no services
Each configuration object has a single instance, even if that instance applies to several hosts. If you apply a contact or escalation to that object it effects that instance of the object and thus every host it's associated with.
Our wizards create one service object per host, this offers more flexibility at the cost of having to make many small(that are script-able) changes later.
Our wizards create one service object per host, this offers more flexibility at the cost of having to make many small(that are script-able) changes later.
Re: Service Escalation lists no services
Thanks for the response. I understand the logic behind a single instance for each configuration object and agree that it gives a much finer level of control. For environments with similar hardware, and basic system services that are all the same it makes a lot of sense for us to reuse the services/monitors by targeting them to a hostgroup. Shouldn't I be able to build escalations for each of these these service objects regardless what hosts are using them? I dont think you understand exactly what I'm seeing in the GUI and just want to make sure we are on the same page.
My problem is when selecting a hostgroup in the services escalation screen I can't see any service objects get listed, so I can't build an escalation. After playing around with this I found that I can only successfully build a service escalation if the service is defined for a single host. If I target a service to a hostgroup, the service escalation form (with the hostgroup selected) will never populate the Services field with that service, so I will never be able to select the service, and never create the escalation. I can get this to work with host groups only when I remove all the hosts but 1. Sort of defeats the purpose of the group. Based on Nagios 3 documentation I believe this should work.
Can you confirm that this in fact isn't working like I expect in the GUI?
Also, It would be really great if you could share a script that can make changes to the nagios services/hosts/configs/etc..., and then update the database so everything is in sync. Is there anything like that available? I was under the impression you couldn't update any of the flat files because they mostly get generated by the DB. Any help on this would be excellent for the next stages of our research.
By the way, this is a great interface for the $$ and I'm very confident this will get approved and become our monitoring solution. So this level of scripting will probably be useful sometime down the road.
Thank you,
My problem is when selecting a hostgroup in the services escalation screen I can't see any service objects get listed, so I can't build an escalation. After playing around with this I found that I can only successfully build a service escalation if the service is defined for a single host. If I target a service to a hostgroup, the service escalation form (with the hostgroup selected) will never populate the Services field with that service, so I will never be able to select the service, and never create the escalation. I can get this to work with host groups only when I remove all the hosts but 1. Sort of defeats the purpose of the group. Based on Nagios 3 documentation I believe this should work.
Can you confirm that this in fact isn't working like I expect in the GUI?
Also, It would be really great if you could share a script that can make changes to the nagios services/hosts/configs/etc..., and then update the database so everything is in sync. Is there anything like that available? I was under the impression you couldn't update any of the flat files because they mostly get generated by the DB. Any help on this would be excellent for the next stages of our research.
By the way, this is a great interface for the $$ and I'm very confident this will get approved and become our monitoring solution. So this level of scripting will probably be useful sometime down the road.
Thank you,
Re: Service Escalation lists no services
It's a catch 22. If we carry options like check command over to each "virtual instance" but not contact information or escalations then that would be a bug. The settings you are looking for is called a template. You save away an example service in the form of a template and then you create a service instance for each host using only the common settings you want to each instance of a service object.
This is where servicegroups come in.
You can call the *NagiosQL engine and feed it config files/objects to be added/updated. We do track bugs in NagiosQL, but there are several UI issues that I'd rather have them spend time on right now. A 90% featured interface is better then a 90% stable one, IMHO.
* There is an example of how to do this in our code. As part of the configuration import and during installation. You drop files into a folder and our scripts carry it off into NagiosQL.
This is where servicegroups come in.
You can call the *NagiosQL engine and feed it config files/objects to be added/updated. We do track bugs in NagiosQL, but there are several UI issues that I'd rather have them spend time on right now. A 90% featured interface is better then a 90% stable one, IMHO.
* There is an example of how to do this in our code. As part of the configuration import and during installation. You drop files into a folder and our scripts carry it off into NagiosQL.
Re: Service Escalation lists no services
I tried to create service templates for each snmp service I want to support on my linux boxes (lets say 5 total services). Then I created a single service, targeted it to my linux host, and inherited all the snmp service templates. I was hoping by inheriting these templates my host would automatically have all 5 services being monitored. It doesn't work this way. Instead, I have that single service that I created, and it appears the inherited services dont really do anything. I guess I can't do this?
Re: Service Escalation lists no services
I'm not sure I follow, perhaps we are just not using the same terms.
http://nagios.sourceforge.net/docs/3_0/
You inherit traits from your parents, but for each treat there can be only one setting... Thus if you get your mother's blue eyes you won't get the brown eyes of your grandfather.
In this example any setting in 1 can only be overridden in devweb1.
To do what you are looking for you would define a template that contains all your SNMP settings like hostgroup, short of the actual check_command. Then define one object for each check_command.
For different escalations as described you would define another set of these objects with a check_command and an over-riding hostgroup using the original template. Use the first instance of this as a template for the rest to avoid having to repeat the hostgroup.
If you had several hundred SNMP mibs and a handful of escalations this would be a problem. In this situation using hostgroups would be essential as would using service groups in your escalations.
http://nagios.sourceforge.net/docs/3_0/
You inherit traits from your parents, but for each treat there can be only one setting... Thus if you get your mother's blue eyes you won't get the brown eyes of your grandfather.
In this example any setting in 1 can only be overridden in devweb1.
To do what you are looking for you would define a template that contains all your SNMP settings like hostgroup, short of the actual check_command. Then define one object for each check_command.
For different escalations as described you would define another set of these objects with a check_command and an over-riding hostgroup using the original template. Use the first instance of this as a template for the rest to avoid having to repeat the hostgroup.
If you had several hundred SNMP mibs and a handful of escalations this would be a problem. In this situation using hostgroups would be essential as would using service groups in your escalations.